Lindsey F. Masson

Statistical approaches for assessing the relative validity of a food-frequency questionnaire: use of correlation coefficients and the kappa statistic.

OBJECTIVE To compare different statistical methods for assessing the relative validity of a self-administered, 150-item, semi-quantitative food-frequency questionnaire (FFQ) with 4-day weighed diet records (WR). DESIGN Subjects completed the Scottish Collaborative Group FFQ and carried out a 4-day WR. Relative agreement between the FFQ and WR for energy-adjusted nutrient intakes was assessed by Pearson and Spearman rank correlation coefficients, the percentages of subjects classified into the same and opposite thirds of intake, and Cohens weighted kappa. SUBJECTS Forty-one men, mean age 36 (range 21-56) years, and 40 women, mean age 33 (range 19-58) years, recruited from different locations in Aberdeen, Scotland. RESULTS Spearman correlation coefficients tended to be lower than Pearson correlation coefficients, and were above 0.5 for 10 of the 27 nutrients in men and 17 of the 27 nutrients in women. For nutrients with Spearman correlation coefficients above 0.5, the percentage of subjects correctly classified into thirds ranged from 39 to 78%, and weighted kappa values ranged from 0.23 to 0.66. CONCLUSIONS Both Spearman correlation coefficients and weighted kappa values are useful in assessing the relative validity of estimates of nutrient intake by FFQs. Spearman correlation coefficients above 0.5, more than 50% of subjects correctly classified and less than 10% of subjects grossly misclassified into thirds, and weighted kappa values above 0.4 are recommended for nutrients of interest in epidemiological studies.

Meal and snacking patterns of school-aged children in Scotland

Background:Current lifestyles and the choice and availability of foods may influence the eating patterns of children. The aim of this study was to investigate the meal and snacking patterns of school-aged children in Scotland.Methods:A sub-sample of 156 children (5–17 years) from the national Survey of Sugar Intake among Children in Scotland completed a 4-day non-weighed diet diary. Meals and snacks were defined using a food-based classification system based on ‘core’ and ‘non-core’ foods. The first eating event containing a solid food item up to and including 0900 hours (1100 hours on weekend days) was defined as breakfast. Frequency of meal and snack consumption was compared between age, sex, body mass index (BMI) and socio-economic sub-groups, between term-time and school holidays and between weekdays and weekend days. Intakes of total fat, saturated fatty acids (SFA) and non-milk extrinsic sugars (NMES) on these days were also compared.Results:Children ate a median of 3.3 meals plus 2.0 snacks per day, which did not vary between age and BMI groups. In all, 83% of children ate breakfast on all 4 days. Boys ate more meals than girls but the number of snacks was similar. Children from lower socio-economic groups tended to eat more meals and fewer snacks. Snacks accounted for 21% of the total daily energy intake, 22% of total fat, 24% of SFA and 39% of NMES intake. Daily intake of energy, total fat, SFA and NMES did not differ between term-time and holidays or weekdays and weekend days.Conclusions:Children tended to follow a traditional pattern of three meals a day, which was consistent between age and BMI subgroups and between school term-time and holidays.

Polymorphisms in the methylenetetrahydrofolate reductase ( MTHFR ) gene, intakes of folate and related B vitamins and colorectal cancer: a case–control study in a population with relatively low folate intake

Folate is key in one-carbon metabolism, disruption of which can interfere with DNA synthesis, repair, and methylation. Efficient one-carbon metabolism requires other B vitamins and the optimal activity of enzymes including 5,10-methylenetetrahydrofolate reductase (MTHFR). We report a population-based case-control study of folate intake, related dietary factors and MTHFR polymorphisms (C677T, A1298C) and colorectal cancer in a population with relatively high colorectal cancer incidence and relatively low folate intake. A total of 264 cases with histologically confirmed incident colorectal cancer and 408 controls participated. There was no clear trend in risk with reported intakes of total, or dietary, folate, riboflavin, vitamin B12 or vitamin B6, nor were there interactions between folate intake and the other B vitamins or alcohol. For C677T, risk decreased with increasing variant alleles (multivariate OR for CT v. CC = 0.77 (95 % CI 0.52, 1.16); OR for TT v. CC = 0.62 (95 % CI 0.31, 1.24)), which, although not statistically significant, was consistent with previous studies. For A1298C, compared with AA subjects, CC subjects had modest, non-significant, reduced risk (multivariate OR = 0.81 (95 % CI 0.45, 1.49)). There were significant interactions between total folate and C677T (P = 0.029) and A1298C (P = 0.025), and total vitamin B6 and both polymorphisms (C677T, P = 0.016; A1298C, P = 0.033), although the patterns observed differed from previous studies. Seen against the setting of low folate intake, the results suggest that the role of folate metabolism in colorectal cancer aetiology may be more complex than previously thought. Investigation of particular folate vitamers (for example, tetrahydrofolate, 5,10-methylenetetrahydrofolate) may help clarify carcinogenesis pathways.

Lycopene intervention reduces inflammation and improves HDL functionality in moderately overweight middle-aged individuals

The management of overweight subjects by interventions aimed at reducing inflammation is highly desirable. To date, observational studies have identified a link between increased dietary antioxidant intake and reduced cardiovascular morbidity. However, direct trial evidence regarding the ability of antioxidants to influence inflammation is lacking. Therefore, this study examined lycopenes ability to lower systemic and high-density lipoprotein (HDL)-associated inflammation in moderately overweight middle-aged subjects. Serum was collected before and after a 12-week intervention from 54 moderately overweight, middle-aged individuals. Subjects were randomised to one of three groups: control diet (<10 mg lycopene/week), lycopene-rich diet (224-350 mg lycopene/week) and lycopene supplement (70 mg lycopene/week). HDL was subfractionated into HDL(2&3) by rapid ultracentrifugation. Compliance was monitored by assessing lycopene concentration in serum and HDL(2&3). Systemic and HDL-associated inflammation was assessed by measuring serum amyloid A (SAA) levels. HDL functionality was determined by monitoring the activities of paraoxonase-1 (PON-1), cholesteryl ester transfer protein (CETP) and lecithin cholesterol acyltransferase (LCAT). Lycopene increased in serum and HDL(2&3) following both lycopene interventions (P<.001, for all), while SAA decreased in serum following the lycopene supplement and in HDL(3) following both lycopene interventions (P<.05 for all). PON-1 activity increased in serum and HDL(2&3) in both lycopene groups (P<.05, for all). Furthermore, the activity of CETP decreased in serum following the lycopene supplement, while the activity of LCAT increased in serum and HDL(3) following both lycopene interventions (P<.05 for all). These results demonstrate that in moderately overweight, middle-aged subjects, increasing lycopene intake leads to changes to HDL(2&3), which we suggest enhanced their antiatherogenic properties. Overall, these results show the heart-protective properties of increased lycopene intake.

The effect of genetic variation on the lipid response to dietary change: recent findings.

Purpose of review Dyslipidaemia is an important risk factor for cardiovascular disease, and can be modified by diet. However, the lipid response to dietary change may be influenced by genetic variation. This review examines recent research (published since August 2003) on the effect of genetic variation on the lipid response to dietary change. Recent findings In 10 reports describing intervention studies and seven reports describing observational studies, the lipid response to diet was modified by polymorphisms within the genes for apoE, apoB, apoCIII, lipoprotein lipase, hepatic lipase, endothelial lipase, the liver fatty acid-binding protein, the β3-adrenergic receptor, adipsin and the peroxisome proliferator-activated receptor γ. The studies varied widely in terms of the number and type of study participants, the composition and duration of the dietary interventions, the nutrients studied and dietary assessment methods used in the observational studies, and the polymorphisms analysed – some of which had not been studied before with regard to the lipid response to diet. Summary The lipid response to dietary change is highly complex. Future studies will have to be large in order to assess the effects of multiple polymorphisms, and will have to control for many factors other than diet. At present, it is premature to recommend the use of genotyping in the design of therapeutic diets. However, such studies may be useful in identifying the mechanisms by which dietary components influence lipid levels. Abbreviations CVD: cardiovascular disease; FFQ: food frequency questionnaire; MUFA: monounsaturated fatty acid; PPAR: peroxisome proliferator-activated receptor; PUFA: polyunsaturated fatty acid; SFA: saturated fatty acid.

Effect of a tomato-rich diet on markers of cardiovascular disease risk in moderately overweight, disease-free, middle-aged adults: a randomized controlled trial

BACKGROUND Cardiovascular disease (CVD) is a major cause of mortality in the United Kingdom. Epidemiologic studies suggest that consumption of tomato-based foods may lower CVD risk. Such potential benefits have been ascribed in part to high concentrations of lycopene in the tomatoes. However, these findings have not yet been validated by comprehensive intervention trials. OBJECTIVE The aim of this study was to conduct a single-blind, randomized controlled intervention trial with healthy middle-aged volunteers to assess whether the consumption of tomato-based foods affects recognized biomarkers of CVD risk. DESIGN After a 4-wk run-in period with a low-tomato diet, 225 volunteers (94 men and 131 women) aged 40-65 y were randomly assigned into 1 of 3 dietary intervention groups and asked to consume a control diet (low in tomato-based foods), a high-tomato-based diet, or a control diet supplemented with lycopene capsules (10 mg/d) for 12 wk. Blood samples were collected at baseline, at 6 wk, and after the intervention and were analyzed for carotenoid and lipid profiles and inflammatory markers. Blood pressure, weight, and arterial stiffness were also measured. Dietary intake was also determined during the intervention. RESULTS None of the systemic markers (inflammatory markers, markers of insulin resistance and sensitivity) changed significantly after the dietary intervention. Moreover, lipid concentrations and arterial stiffness were also unaffected by the interventions. CONCLUSION These data indicate that a relatively high daily consumption of tomato-based products (equivalent to 32-50 mg lycopene/d) or lycopene supplements (10 mg/d) is ineffective at reducing conventional CVD risk markers in moderately overweight, healthy, middle-aged individuals. This trial was registered at isrctn.org as ISRCTN34203810.

Colorectal cancer and genetic polymorphisms of CYP1A1, GSTM1 and GSTT1: A case‐control study in the Grampian region of Scotland

Cytochrome P‐450 CYP1A1 is involved in the metabolism of polycyclic aromatic hydrocarbons (PAHs) that are derived from meat intake and tobacco smoking. Expression of the CYP1A1 gene is induced by compounds present in cruciferous vegetables. The glutathione S‐transferases play a central role in the detoxification of carcinogens, including PAHs. We investigated the association between colorectal cancer and three variants (CYP1A1*2A, CYP1A1*2C, CYP1A1*4) of the CYP1A1 gene, and homozygosity for the null deletion of the GSTM1 and GSTT1 genes, and the joint effects of these genotypes and smoking, meat intake and intake of green leafy vegetables in a population‐based study of 264 cases and 408 controls in Northeast Scotland. There was an inverse association with the CYP1A1*4 (m4) variant (OR 0.3, 95% CI 0.13–0.70). The OR for the CYP1A1*2C (m2) variant was 1.3 (95% CI 0.59–2.91), which is similar to a combined estimate for previous studies (OR 1.2, 95% CI 0.95–1.41). We observed no association with the CYP1A1*2A (m1) variant, or the GSTM1 and GSTT1 polymorphisms. Significant interactions between all 3 CYP1A1 variants and meat intake, and between the m1 and m2 variants and intake of green leafy vegetables, were observed. There was no evidence of interaction between CYP1A1 and smoking, and no evidence of interaction between the GSTM1 or GSTT1 polymorphisms and smoking, meat intake, green leafy vegetable intake, CYP1A1 variants or each other.

A novel online Food Recall Checklist for use in an undergraduate student population: a comparison with diet diaries

BackgroundUniversity students are commonly overlooked when diet of populations is measured and there is a lack of comprehensive dietary assessment in whole university student populations. To measure diet of undergraduate students, a new online 121-item Food Recall Checklist (FoRC) was designed as an alternative to a non-weighed record (food diary). This article reports the comparison between the new dietary assessment method (FoRC) and the food diary as a measure of energy (kJ), fat (g), Non-Starch Polysaccharide (NSP) (g), fruit and vegetables (g), breakfast cereal (g) and bread (g) and alcohol (units) intake.MethodsFifty-three students at the University of Aberdeen completed four days of FoRC then four days food diary. Median agreement and correlation between the two methods was assessed for foods and nutrients using the Spearmans rank correlation co-efficient and the Wilcoxon signed ranks test. Agreement between FoRC and food diary was assessed using the Bland-Altman method.ResultsThe mean time taken to complete FoRC for one day was 7.4 minutes. Intakes of fat (g and % food energy), NSP and bread were similar between FoRC and the food diary. Median energy intake was 8185 kJ in the food diary and 8007 kJ in FoRC. However, FoRC recorded significantly lower intakes of energy and alcohol and significantly higher intakes of fruit and vegetables and breakfast cereal compared with the food diary. There was considerable variation in agreement between methods at the individual level. For all variables except alcohol and percentage energy from fat, correlation co-efficients were statistically significant and greater than 0.5.ConclusionAt the group level, four days of FoRC showed good median agreement with the food diary and there was high correlation between methods for most foods and nutrients. This suggests that this novel method of assessing diet can provide a useful alternative for assessing group mean intakes but that individual intakes may need to be interpreted with care.

Sugar intake and dental decay: results from a national survey of children in Scotland.

The aim of this analysis was to investigate the strength of the association between sugar intake and treatment for dental decay in children in Scotland, and the impact of tooth brushing frequency on this association. The Survey of Sugar Intake among Children in Scotland was carried out in 2006 in those aged 3-17 years. Diet was assessed using the Scottish Collaborative Group FFQ, and interviews were carried out by trained fieldworkers who asked about dental health. A total of 1700 interviews were carried out, and 1512 FFQ were returned. Of the children, 56% had received treatment for decay (fillings or teeth removed due to decay). Intake of non-milk extrinsic sugars (NMES), but not total sugar, increased the risk of having had treatment for decay: adjusted OR 1.84 (95% CI 1.28, 2.64) for the highest ( ≥ 20.0% food energy) v. lowest ( ≤ 14.8% food energy) tertile of NMES intake. This raised risk remained in children who reported brushing their teeth at least twice a day. Compared with children who reported brushing their teeth at least twice a day and were in the lowest tertile of NMES intake, children who reported brushing their teeth once a day or less and were in the highest tertile of NMES intake were over three times more likely to have received treatment for decay (adjusted OR 3.39, 95% CI 1.97, 5.82). In order to improve dental health in children in Scotland, dental health strategies must continue to stress the importance of both reduced NMES intake and good oral hygiene.

Cardiovascular benefits of lycopene: fantasy or reality?

Epidemiological evidence indicates that high consumption of tomatoes and tomato-based products reduces the risk of chronic diseases such as CVD and cancer. Such potential benefits are often ascribed to high concentrations of lycopene present in tomato products. Mainly from the results of in vitro studies, potential biological mechanisms by which carotenoids could protect against heart disease and cancer have been suggested. These include cholesterol reduction, inhibition of oxidation processes, modulation of inflammatory markers, enhanced intercellular communication, inhibition of tumourigenesis and induction of apoptosis, metabolism to retinoids and antiangiogenic effects. However, with regard to CVD, results from intervention studies gave mixed results. Over fifty human intervention trials with lycopene supplements or tomato-based products have been conducted to date, the majority being underpowered. Many showed some beneficial effects but mostly on non-established cardiovascular risk markers such as lipid peroxidation, DNA oxidative damage, platelet activation and inflammatory markers. Only a few studies showed improvement in lipid profiles, C reactive protein and blood pressure. However, recent findings indicate that lycopene could exert cardiovascular protection by lowering HDL-associated inflammation, as well as by modulating HDL functionality towards an antiatherogenic phenotype. Furthermore, in vitro studies indicate that lycopene could modulate T lymphocyte activity, which would also inhibit atherogenic processes and confer cardiovascular protection. These findings also suggest that HDL functionality deserves further consideration as a potential early marker for CVD risk, modifiable by dietary factors such as lycopene.